Anthropic Alleges Alibaba's Unauthorized Access to AI Model Claude

A recent accusation has been made by Anthropic, a US-based company, against Alibaba Group, a Chinese conglomerate, for allegedly attempting to gain unauthorized access to Anthropic's artificial intelligence model, Claude. According to Anthropic, Alibaba utilized approximately 25,000 fake profiles to interact with Claude between April and June. This alleged campaign, reportedly carried out by operators linked to Alibaba's AI lab, Qwen, targeted the most valuable aspects of Claude, including software development and agent thinking. Anthropic claims that this tactic, known as data distillation, is used to illicitly acquire advanced American AI technologies, replicate them, and then pass them off as their own, all without incurring the costs associated with research and development.

The company has urged Washington to take action against firms that employ such methods to gather data for creating their own AI models. This incident highlights the growing concerns surrounding AI model security and the need for stricter regulations to prevent unauthorized access and data theft.

Source: RIA Novosti